[gmx-users] add parameter

[Justin Lemkul]

On 2/11/16 12:03 PM, Malihe Hasanzadeh wrote:
> Hi
> I have three water molecules and one new residue that I want to add their
> parameters to force field Amber99sb. One of water molecules has only one
> hydrogen, and since I want to prevent from replacing with SOL in solvation
> step, So I define them as new residue in .rtp file:
>
> [ INN ]
>   [ atoms ]
>     H1    H           0.384620     1
>      O    O          -0.671930     2
>     H2    H           0.384620     3
>
>
> [ INT ]
>   [ atoms ]
>     H1    H           0.286310     1
>      O    O          -0.341910     2
>     H2    H           0.286310     3
>
>
> [ INB ]
>   [ atoms ]
>     H1    H           0.477860     1
>      O    O          -1.092580     2
> Also define them in .dat file as Protein  and in .hdb file as below:
>

Defining these residues as protein is not necessary.

> INN    2
> 1    7    H1    O
> 1    7    H2    O
> INT    2
> 1    7    H1    O
> 1    7    H2    O
> INB    1
> 1    7    H1    O
>
> But when I start MD simulation(gromacs 5.0.4) in step of mdrun ...em,
> gives segmentation fault! when I remove their parameters in .hdb file, it
> is done but without any hydrogen for them, just three oxygen atoms in
> em.gro!
>

This is because you would have had to invoke -missing with pdb2gmx and yield 
something totally useless.

Immediate seg faults are indicative of complete instability.  Note that none of 
your .rtp entries have [bonds] defined, so you probably just have particles very 
close to one another (e.g. as they would be in a real molecule) but no covalent 
attachment, so they likely just fly apart.

> Also I have another problem with that new residue, this residue is like a
> Lys but has two carbonyl groups binded to extra carbon instead of amino
> group in Lys. So I used Lys parameters for that and add some bonds and
> atoms to it:
>
> [ NEW ] ;
>   [ atoms ]
>      N       N
>      H       H
>     CA       CT
>     HA        H1
>     CB       CT
>    HB1       HC
>    HB2       HC
>     CG      CT
>    HG1      H
>    HG2      H
>     CD      CT
>    HD1      HC
>    HD2      HC
>     CE      CT
>    HE1      HP
>    HE2      HP
>     NZ      N
>     HZ      H
>     CX      C
>    OQ1     O
>    OQ2     O
>      C      C
>      O      O
>   [ bonds ]
>       N     H
>       N    CA
>      CA    HA
>      CA    CB
>      CA     C
>      CB   HB1
>      CB   HB2
>      CB    CG
>      CG   HG1
>      CG   HG2
>      CG    CD
>      CD   HD1
>      CD   HD2
>      CD    CE
>      CE   HE1
>      CE   HE2
>      CE    NZ
>      NZ    HZ
>      NZ    CX
>       C     O
>      CX   OQ1
>      CX   OQ2
>   [ impropers ]
>      -C    CA     N     H
>      CA    +N     C     O
> when I added only atoms parameters in .rtp file, finally in em.gro file
> there wasn't any bond between atoms of new residue and when I added bond

As you should expect.  See above.  [bonds] is a mandatory directive (which is 
also stated clearly in the manual).

> parameters also, in grompp gives 10 error about "default angel and
> dihedral"! I find the number of angels in .top file and added them to
> ffbonded.itp, but anything have no change.

You're adding moieties that don't exist in the force field, unless you have 
parameters from some external source (e.g. something published or otherwise 
available in a force field repository).

-Justin

-- 
==================================================

Justin A. Lemkul, Ph.D.
Ruth L. Kirschstein NRSA Postdoctoral Fellow

Department of Pharmaceutical Sciences
School of Pharmacy
Health Sciences Facility II, Room 629
University of Maryland, Baltimore
20 Penn St.
Baltimore, MD 21201

jalemkul at outerbanks.umaryland.edu | (410) 706-7441
http://mackerell.umaryland.edu/~jalemkul

==================================================